惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

Y
Y Combinator Blog
V
V2EX
Jina AI
Jina AI
爱范儿
爱范儿
M
MIT News - Artificial intelligence
量子位
L
LangChain Blog
Google DeepMind News
Google DeepMind News
酷 壳 – CoolShell
酷 壳 – CoolShell
罗磊的独立博客
腾讯CDC
MongoDB | Blog
MongoDB | Blog
P
Proofpoint News Feed
宝玉的分享
宝玉的分享
OSCHINA 社区最新新闻
OSCHINA 社区最新新闻
Last Week in AI
Last Week in AI
H
Hackread – Cybersecurity News, Data Breaches, AI and More
F
Fortinet All Blogs
The GitHub Blog
The GitHub Blog
Engineering at Meta
Engineering at Meta
博客园 - 聂微东
让小产品的独立变现更简单 - ezindie.com
让小产品的独立变现更简单 - ezindie.com
Vercel News
Vercel News
T
The Blog of Author Tim Ferriss

DEV Community

Authentication Security Deep Dive: From Brute Force to Salted Hashing (With Java Examples) Why AI Systems Don’t Fail — They Drift Spilling beans for how i learn for exam😁"Reinforcement Learning Cheat Sheet" I Replaced Chrome with Safari for AI Browser Automation. Here's What Broke (and What Finally Worked) How Python Borrows Other People's Work The $40 Architecture: Processing 1 Billion API Requests with 99.99% Uptime Vibe Coding: A Workflow Guide (From Zero to SaaS) Most webhook security guides protect the wrong side. The scary part is delivery. Headless CMS for TanStack Start: Build a Blog with Cosmic EU Age Verification App "Hacked in 2 Minutes" — What Actually Happened Comfy Cloud’s delete function does not actually remove files Running AI Models on GPU Cloud Servers: A Beginner Guide Event-driven media intelligence with AWS Step Functions and Bedrock I scored 500 AI prompts across 8 quality dimensions — here's what broke How to Call Google Gemini API from Next.js (Free Tier, No Backend Needed) The Portal Protocol: Reclaiming Human Connection in the Age of AI How to Fix Your Team's Scattered Knowledge Problem With a Self-Hosted Forum Intro to tc Cloud Functors: A Graph-First Mental Model for the Modern Cloud Designing Multi-Tenant Backends With Both Ownership and Team Access I Built a Neumorphic CSS Library with 77+ Components — Here's What I Learned PostgreSQL Performance Optimization: Why Connection Pooling Is Critical at Scale Cómo construí un SaaS multi-rubro para gestionar expensas en Argentina con FastAPI + Vue 3 🚀 I Built an Ethical Hacking Scanner Tool – Open Source Project I Replaced /usage and /context in Claude Code With a Single Statusline A Pythonic Way to Handle Emails (IMAP/SMTP) with Auto-Discovery and AI-Ready Design I Collected 8.9 Million Polymarket Price Points — Here's What I Found About How Markets Really Move EcoTrack AI — Carbon Footprint Tracker & Dashboard Everyone's Using AI. No One Agrees How. 5 self-hosted ebook managers worth trying in 2026 Building Your First AI Agent with LangChain: From Chatbot to Autonomous Assistant
We Need a CatRun for the AI Era
taekim34 · 2026-05-05 · via DEV Community

A 16-pixel hero in your macOS menu bar. Watches LLM traffic. That's it.

RunCat told us the CPU was busy. Nothing tells us the agent is.

You remember RunCat — the kitten in your menu bar that runs faster when your CPU is busy. Almost a decade old. Adorable. Useful. Asks nothing of you.

AI-native development needs the same thing for a different signal. Not CPU. Agent traffic. Is there a live LLM request flowing right now, or is everything quiet?

That's why I built AgentRunner.

We need a CatRun for the AI era. So I made one.

A 16-pixel hero in your macOS menu bar. Runs when your agent's actually working. Idle when it isn't. That's the whole UI.

AgentRunner demo — 23-second walkthrough

Seven things it's built around

1. The menu bar is where you already glance. Same place as the clock. No extra tab, no extra window, no "I'll open the dashboard later."

2. Below the noise floor. <1% CPU, ~20MB RAM. Native SwiftUI. A monitor that becomes its own monitoring problem is a joke.

3. Flashy "live agent dashboards" don't last. Animated traffic, live token deltas, color-coded latency heatmaps — fun for a week, closed and forgotten by the next sprint. CatRun ran for a decade because it asked you nothing. Same spirit here.

4. Detailed analysis belongs in a different tool. Token spend, cache misses, run history — that needs report depth. That's what cc-token-saver is for, and it gets its own post next. AgentRunner = glance. cc-token-saver = report. Don't make one app try to be both.

5. Vendor-neutral by design. It watches LLM traffic, not Claude traffic. Claude Code, Codex, Cursor, Windsurf, local LLaMA via Ollama, any agent loop hitting a model endpoint over HTTPS. No API key, no per-vendor SDK.

6. Local-only. Zero telemetry. Detection happens on your machine. The app does not phone home. No analytics SDK, no event ping. An agent monitor that ships your data anywhere doesn't deserve trust.

7. Idle vs Active. Binary. That's the entire UI. CatRun gave us a kitten that ran when CPU spiked. AgentRunner gives you a 16-pixel hero that runs when LLM traffic flows. Same spirit. Useful. Small. Invisible until you glance at it.

Get it

cc-token-saver post: coming next.